visualization
Qiuyi Lu
12/14/2019
Heat Map:
We can use heat map to see correlations among variables.
It’s useful for selecting features when building models. For linear regression model, it can also help check multicollinearity.
Many bio companies use heat map to see the gene similarity among different species and could cluster species.
## mpg cyl disp hp drat wt qsec
## mpg 1 -0.852162 -0.8475514 -0.7761684 0.6811719 -0.8676594 0.41868403
## cyl NA 1.000000 0.9020329 0.8324475 -0.6999381 0.7824958 -0.59124207
## disp NA NA 1.0000000 0.7909486 -0.7102139 0.8879799 -0.43369788
## hp NA NA NA 1.0000000 -0.4487591 0.6587479 -0.70822339
## drat NA NA NA NA 1.0000000 -0.7124406 0.09120476
## wt NA NA NA NA NA 1.0000000 -0.17471588
## qsec NA NA NA NA NA NA 1.00000000
## vs NA NA NA NA NA NA NA
## am NA NA NA NA NA NA NA
## gear NA NA NA NA NA NA NA
## carb NA NA NA NA NA NA NA
## vs am gear carb
## mpg 0.6640389 0.5998324 0.4802848 -0.55092507
## cyl -0.8108118 -0.5226070 -0.4926866 0.52698829
## disp -0.7104159 -0.5912270 -0.5555692 0.39497686
## hp -0.7230967 -0.2432043 -0.1257043 0.74981247
## drat 0.4402785 0.7127111 0.6996101 -0.09078980
## wt -0.5549157 -0.6924953 -0.5832870 0.42760594
## qsec 0.7445354 -0.2298609 -0.2126822 -0.65624923
## vs 1.0000000 0.1683451 0.2060233 -0.56960714
## am NA 1.0000000 0.7940588 0.05753435
## gear NA NA 1.0000000 0.27407284
## carb NA NA NA 1.00000000
Geographic map
We can use map to find the popularity of some places. If we want to set a new station, we can check where do people take a lot of activities and set points there. Rideshare companies can use it to better allocate their drivers and pricing. Travelling/local service companies can recommend the place for users.
## # A tibble: 1 x 5
## start.station.id lat long name n.trips
## <int> <dbl> <dbl> <chr> <int>
## 1 521 40.8 -74.0 8 Ave & W 31 St 14498
World map:
US MAP
